From f85c43f00ca294a48a7849bf7fa103fcbc8a6521 Mon Sep 17 00:00:00 2001
From: Ryo Neyama <neyama@toyota-tokyo.tech>
Date: Thu, 14 Nov 2024 08:20:14 +0000
Subject: [PATCH] clear the cached tables before each query execution

---
 runners/datafusion-comet/tpcbench.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/runners/datafusion-comet/tpcbench.py b/runners/datafusion-comet/tpcbench.py
index ed11938..3ae958c 100644
--- a/runners/datafusion-comet/tpcbench.py
+++ b/runners/datafusion-comet/tpcbench.py
@@ -61,6 +61,7 @@ def main(benchmark: str, data_path: str, query_path: str, iterations: int, outpu
         print(f"Starting iteration {iteration} of {iterations}")
 
         for query in range(1, num_queries+1):
+            spark.catalog.clearCache()
             spark.sparkContext.setJobDescription(f"{benchmark} q{query}")
 
             # read text file